On Tue, Jul 15, 2014 at 10:48 AM, Miika Turkia <[email protected]> wrote:
> On Tue, Jul 15, 2014 at 10:26 AM, Robert Helling <[email protected]> > wrote: > >> >> On 15.07.2014, at 07:14, Miika Turkia <[email protected]> wrote: >> >> There is something wrong currently with the calculation of tissue >> saturation. It looks like the earlier dive is getting cumulative saturation >> from the later dive. Clearly visible when you open the XML and switch >> between the dives. There is also a spike on the start of the earlier dive >> as show on screenshot. >> >> This is with the beta release (and master). >> >> I have added a test XML to demonstrate the issue. dives/test20.xml with >> the same dive repeated again the next day. >> >> miika >> >> >> Miika, >> >> TL;DR: Works as designed. >> >> I believe this is fine. It is the outcome of the problem with “negative >> slope” that we had a few weeks ago. The point is that gradient factors are >> not really well defined during the early descent phase of the dive (as >> GFlow is defined to apply at the depth of the first stop but there is not >> really a first stop when you just entered the water). >> >> For a longer explanation, see my commit message for >> e38a473a4d9119b86e5484a03dbf2cafdb8d145c >> (and also some posts to the mailing list I wrote in connection with it) >> >> If you think, we need to explain this to our users, I could try to write >> something for the manual (or alternatively: a web page to which the manual >> could point). But unlikely this week as it’s super busy at work. >> > > The main problem is that the accumulated tissue saturation is shown on the > dive on day one. If you compare the tissue graph on these two dives, the > day two dive shows less saturation on the early part of the dive (around 5 > minutes) than the dive on the previous day. The visual glitch was just an > extra oddity I noticed. > OK, now that I play with the planner, it really seems that this is working properly. I first thought that your explanation was only about the visual glitch on the during the first few seconds of the dive, but I suppose it is about the whole issue I am seeing. It really seems weird with the same dives copied to different days, but I don't think it should be as obvious for people who are not manipulating the log files manually. Especially with real dives this should not be that obvious... miika
_______________________________________________ subsurface mailing list [email protected] http://lists.hohndel.org/cgi-bin/mailman/listinfo/subsurface
